Urinary incontinence (UI) is highly prevalent, affecting women of all ages with significant impact on quality of life. Prevalence of stress UI (i.e. loss of urine on exertion) peaks in midlife, and then declines thereafter. The adult female prevalence is 13%. Prevalence of urge (i.e. loss of urine with a strong desire to urinate) and mixed (stress and urge) UI increase with age with prevalence estimates of 5% and 11%, respectively. Importantly, prevalence of mixed UI is far more common than expected if pure stress and pure urge UI are assumed to be independent. The expected (E) co-occurrence is 0.65% (0.13*0.05 = 0.0065), compared to an observed (O) co-occurrence of 11%. The purpose of the application is to understand why stress and urge UI co-occur substantially more often than expected. We propose to test the validity of three possible models. The liability model posits that the presence of one UI subtype increases the likelihood for onset of the other, or that remission of either subtype is lowest in those with mixed UI. An alternative explanation is that mixed UI is associated with more persistent and frequent symptoms than pure stress or urge UI. In this severity model, women with mixed UI are more likely to exhibit UI symptoms on any given day (i.e., more likely to be detected at cross-section as an active case) than their pure stress or pure urge counterparts, because they inherently have a more severe condition. A third explanation is that risk factors common to two different conditions can lead to co-occurrence more often than expected. Risk factors common to stress and urge UI include obesity, parity, hormone replacement therapy, prolapse and others. We propose a four- year longitudinal study to: 1) Determine if the individual or composite effect of incidence and remission of UI subtypes explains the excess prevalence of mixed UI; and 2) Determine if the expression of incontinence symptoms is more frequent and persistent for mixed UI compared to either pure stress or pure urge UI cases. Within the longitudinal study we also propose a nested case-control study with a comprehensive evaluation of risk factors to determine if risk factors common to both stress and urge UI explain the excess co-occurrence of these two UI subtypes. Understanding the relative influence of each model on prevalence of UI subtypes will help to explain etiology of UI subtypes and prognosis, information that we believe is directly relevant to clinical care. PUBLIC HEALTH REVELANCE The proposed project will increase our knowledge of the natural history of urinary incontinence in women including incidence, remission and transition rates among stress, urge and mixed urinary incontinence cases. Moreover, it will result in a better understanding of the relation of risk factors to the evolution of incontinence subtypes.